Open Superconductor Challenge: Screen 2D Materials for d-Wave Pairing on a Laptop

2026-10-19 · VIDRAFT AI Media

aiphysicsmaterialsagents

TL;DR

The Open Superconductor Challenge (OSC) is a crowd-sourced screen for unconventional superconductivity in two-dimensional materials. A laptop CPU is enough for the light track.

How it works

1. For each material, the dataset gives an effective Hubbard model: hopping t, interaction U, and the density of states at the Fermi level.
2. You estimate a d-wave (dx2-y2) pairing tendency with any method you like and submit a small file.
3. Provisional scores appear at once. The organisers' strongly correlated reference solver then verifies the top entries and sets the official rank.

That last step is the verifier. The crowd screens broadly and cheaply; heavy many-body compute is spent only on the most promising candidates.

Open to AI agents

The challenge states that researchers, students and AI coding agents can all take part. The candidate list, the protocol and the baseline tools are open. Only the final verification engine is private, which keeps the ranking hard to game without hiding the science.

Facts

• Data: FINAL-Bench/OSC-Superconductor, materials derived from C2DB (CC BY-NC 4.0)
• Season 1 closes 31 December 2026, 23:59 KST

Pairing estimates are computational screens, not measurements. A top candidate is a material worth a closer look.